Previous Article | Next Article ![]()
Journal of Virology, December 2003, p. 13335-13347, Vol. 77, No. 24
0022-538X/03/$08.00+0 DOI: 10.1128/JVI.77.24.13335-13347.2003
Copyright © 2003, American
Society for
Microbiology. All Rights Reserved.
Hiroko Amano,2 Yoshiaki Ueda,2 Jing Qin,1 Tatsuo Miyamura,2 Tetsuro Suzuki,2 Xing Li,3 John W. Barrett,1 and Grant McFadden1,4*
BioTherapeutics Research Group, Robarts Research Institute, London, Ontario, Canada N6G 2V4,1 Department of Microbiology and Immunology, University of Western Ontario, London, Ontario, Canada N6A 5C1,4 Laboratory of Tumor Viruses, National Institute of Infectious Diseases, Tokyo 162-8640, Japan,2 Viron Therapeutics Inc., London, Ontario, Canada N5G 2V43
Received 15 July 2003/ Accepted 4 September 2003
|
|
|---|
|
|
|---|
The genome sizes of published chordopoxviruses vary from 145,000 bp for Yaba-like disease virus (YLDV) (15) up to 288,000 bp for fowlpox virus (2) and possess between 151 and 260 assigned open reading frames (ORFs). Complete genomic sequences of representative members from seven of the eight Chordopoxvirus genera have now been published, including orthopoxviruses (vaccinia virus strain Copenhagen [11], modified vaccinia virus strain Ankara [6], variola virus strain Bangladesh [16], variola virus strain India [24], variola virus strain Garcia [25], camelpox virus [1], and monkeypox virus [26]), capripoxviruses (lumpy skin disease virus [LSDV] [29],goatpox virus, and sheeppox virus [30]), leporipoxviruses (myxoma virus [8] and Shope fibroma virus [31]), suipoxviruses (swinepox virus [SPV] [3]), molluscipoxvirus (molluscum contagiosum virus [23]), avipoxviruses (fowlpox [2]), and yatapoxviruses (Yaba-like disease virus [YLDV] [15]).
The Yatapoxvirus genus of poxviruses is comprised of three virus isolates: YLDV, Tanapox virus (TPV), and Yaba monkey tumor virus (YMTV) (14). The yatapoxviruses have a narrow host range, infecting only primates, including humans. Several pieces of data suggest that TPV and YLDV may be different strains of the same virus. For example, TPV and YLDV produce a clinically indistinguishable disease, which includes a mild fever and epidermal lesions (10, 17), and the published genomic sequence of YLDV is more than 98.6% identical with the 8,300 bases of TPV sequence entered into the public database (GenBank accession no. AY253325, AF245394, and AF153912) (15). This level of sequence identity is comparable to different strains of vaccinia virus and suggests that YLDV and TPV should be considered the monkey and human versions, respectively, of the same virus.
YMTV was originally characterized to be the agent responsible for subcutaneous tumors in a rhesus monkey colony occurring in 1956 in Yaba, Nigeria (7). YMTV is one of the few poxviruses that induce substantial tumor formation upon infection (5, 12, 20, 27). In rhesus monkeys infected with YMTV, the tumors are thought to be derived from histiocytes that migrate to the site of infection. The histiocytes become infected and begin to rapidly proliferate, become multinucleated, and eventually form a polyclonal tumor (27). However, the tumors generally do not become invasive and spontaneously regress, presumably when either viral cytopathic effect kills the infected cells or cell-mediated antiviral immunity becomes sufficiently effective to clear the infection (12, 27).
The complete genomic sequence of YLDV was recently published, and a number of novel ORFs not found in other chordopoxviruses were identified (15). As well, despite the fact that the noncoding regions between ORFs in most poxviruses are typically only a few nucleotides, there were multiple identified inter-ORF regions of 200 or more nucleotides in YLDV. Typically, the minimum size for a poxvirus ORF is arbitrarily set (e.g., 30 amino acids for SPV, LSDV, molluscum contagiosum virus, and fowlpox virus [2, 3, 23, 29]; 50 amino acids for myxoma virus [8]; and 60 amino acids for YLDV [15]). If bona fide ORFs were indeed located within these assigned YLDV noncoding regions, then one would predict that these ORFs might be highly conserved between YLDV and YMTV. Therefore, in an effort to understand the clinical differences between YLDV and YMTV and to provide a closely related sequence to YLDV for a comparative genomic approach, we sequenced the genome of the YMTV and provide a comparative genomic analysis of the Yatapoxvirus genus.
|
|
|---|
Isolation and sequencing of YMTV genomic fragments. YMTV genomic DNA was isolated from infected CV1 cells and was subjected to restriction enzyme digestion with PstI, BamHI, SalI, XbaI, or EcoRI. The digested DNA was cloned into pUC19 or pBR322 vectors and sequenced by the dideoxy sequencing method (21). The remainder of the YMTV genomic sequence was cloned using overlapping PCR. Briefly, PCR was carried out using Taq polymerase, YMTV genomic DNA, and PCR primers based on the corresponding sequence of YLDV (15). The resulting PCR products were cloned into pGEMT-easy (Promega, Madison, Wis.) and were sequenced by the London Regional Genomics Centre DNA Sequencing Facility using an Applied Biosystems (Foster City, Calif.) ABI Prism 377 DNA sequencer and Big Dye terminators (Applied Biosystems). Some of the YMTV sequence was previously submitted to GenBank (accession no. AY253324, AB025319, AB018404, and AB015885).
Sequence analysis. The sequence data were assembled using Sequencher 3.0, and ORFs were identified using MacVector 6.5.3 (Oxford Molecular Ltd.).
Cloning a conserved sequence from myxoma virus upstream of an enhanced GFP cassette. PCR was carried out using Taq polymerase; plasmid DNA pEGFP-N1 (Clontech, Palo Alto, Calif.); the reverse PCR primer 5' TTACGCCTTAAGATACATTG 3', which corresponds to the 3' end of the green fluorescent protein (GFP), and the forward PCR primers (with the start codon of GFP in boldface type) 5' TCGCCACCATGGTGAGCAAG 3' (PCR-GFP), 5' TTTATTTATGTTATTAGCTAGGATTTATGTTTCATTTTTTACTCGCCACCATGGTGAGCAAG 3' (PCR-R-GFP), and 5' GTAAAAAATGAAACATAAATCCTAGCTAATAACATAAATAAATCGCCACCATGGTGAGCAAG 3' (PCR-L-GFP). The resulting PCR products were cloned into pGEMT-easy (Promega) and designated GFP, R-GFP, and L-GFP.
Expression of GFP cassette in BGMK cells. Twelve-well dishes of BGMK cells approximately 90% confluent growing in minimum essential medium-5% fetal bovine serum were either infected with myxoma virus at a multiplicity of infection of 10 or mock infected. The cells were incubated at 37°C for 2 h, and this was followed by transfection with GFP, R-GFP, or L-GFP plasmid DNA using Lipofectamine Plus (Invitrogen, Burlington, Ontario, Canada) per the manufacturer's protocol. The cells were subsequently incubated at 37°C for 48 h. Cells expressing the GFP construct were detected using a fluorescence microscope.
Nucleotide sequence accession number. Sequence data from this article have been deposited in GenBank under accession number AY386371.
RESULTS
Genome structure of YMTV.
The genome of YMTV was sequenced
through the subcloning of genomic fragments into plasmid vectors, and
clones were individually sequenced. In addition, regions of the genome
not represented in the cloned fragments were isolated using PCR, and a
minimum of three independent PCR products for each primer set were
sequenced. After assembling the sequence files, a single continuous
sequence of 134,721 bases was generated, making YMTV the smallest
poxvirus genome yet sequenced. This deduced sequence lacks the terminal
hairpin region, but evidence suggests that all the coding ORFs have
been fully sequenced and only the very extreme hairpin termini of the
genome were not included. In particular, the putative YMTV concatemer
resolution sequence was obtained, which is typically found very close
to the molecular hairpin loop at the termini
(18). Published reports
also confirm that the YMTV genome size is indeed approximately 135,000
bases (4).
The YMTV genome has an A+T content of 70.2% and encodes at least 140 ORFs (Table 1; Fig. 1), of which 139 are single copies and 1 is repeated in each copy of the TIR. In comparison, YLDV has been assigned 151 ORFs (15). YMTV and YLDV is closely related viruses with approximately 75% identity between the viruses at the nucleotide level overall, which is typical for chordopoxvirus members from a single genus. Furthermore, all the ORFs identified in YMTV have a corresponding ortholog in YLDV, but YMTV has lost 13 ORFs that are present in YLDV (Table 2), which accounts for the 10 kb of sequence loss in YMTV. Since YMTV and YLDV are so similar, and to avoid unnecessary confusion, we have adopted the proposed YLDV nomenclature (15) for naming orthologous YMTV ORFs.
|
View this table: [in a new window] |
TABLE 1. YMTV
ORFs
|
![]() View larger version (18K): [in a new window] |
FIG. 1. YMTV
genomic map. The assigned ORFs of YMTV are displayed, with an arrow
indicating the direction of transcription. Each ORF is colored to
indicate a general functional category. In addition, the black arrows
above the ORFs at either end of the genome indicate the
TIR.
|
|
View this table: [in a new window] |
TABLE 2. Immune
evasion ORFs and ORFs absent from YMTV but present in YLDV
|
![]() View larger version (36K): [in a new window] |
FIG. 2. YMTV
and YLDV each contain an apparent pseudogene within the noncoding
region of the termini. An alignment of the assigned SPV002 ORF
(3) with a portion of the
noncoding region of the YMTV and YLDV termini is
shown.
|
![]() View larger version (18K): [in a new window] |
FIG. 3. Comparative
analysis of the noncoding region between YMTV22L and YMTV24L. (a)
Schematic arrangement of ORFs in the region of the 23.5L orthologs in
YMTV and YLDV. A 42-base conserved sequence (CS) is shown as a black
box. (b) Alignment of orthologs of 23.5L in YMTV, YLDV, LSDV (LSDV023),
myxoma virus (M018L), vaccinia virus (vv F8L), fowlpox (FPV113), and
molluscum contagiosum virus (MC014.1L). The SPV (SPV20.5) sequence is
also shown, although it lacks a start codon in the published sequence
(3).
|
Unusual conserved promoter-like sequence found in yatapox, suipox, capripox, and leporipox viruses.
The identification
of the 212-bp gene YMTV23.5L greatly reduced the amount of assigned
noncoding sequence in the region between 22L and 24L. Nevertheless,
when we continued our analysis of the noncoding sequence in this region
between the ORFs 23.5L and 24L in YMTV and YLDV, we noticed a striking
42-bp sequence that was 100% identical between YMTV and YLDV
(Fig.
4a).
![]() View larger version (34K): [in a new window] |
FIG. 4. Analysis
of the conserved promoter-like sequence between YMTV23.5L and YMTV24L.
(a) Alignment of a conserved sequence found between orthologs of
YMTV23.5L and YMTV24L. The sequence within the red boxes labeled
"A" and "B" represents the 9-bp repeat.
The numbers above the grey box indicate the nucleotide positions. (b)
Schematic of the orientation of the promoterless GFP with respect to
the orientation of the cloned myxoma virus conserved sequence. R-GFP
contains the conserved sequence from bases 2 to 41. L-GFP contains the
reverse complement of the conserved sequence. The red boxes show the
location of the two repeats in panel a. (c) Cells were either mock
infected or infected with myxoma virus and subsequently transfected
with either a promoterless GFP, R-GFP, or L-GFP construct. Forty-eight
hours postinfection, the cells were visualized using a fluorescence
microscope.
|
Analysis of the sequence identified two 9-bp repeats separated by 10 bases (Fig. 4a). Since one turn of the DNA double helix is 10.4 bp, this suggests that the two repeats are registered on the same face of the DNA molecule. One possible function for this type of sequence arrangement is the binding of transcription factors to the DNA sequence, and indeed the sequence does resemble a tandem repeat of a canonical poxvirus late promoter (9). To test whether the conserved sequence might function as a viral promoter element, we inserted the conserved 42-bp sequence (derived from myxoma virus) in either the forward (R-GFP) or the reverse complement (L-GFP) orientations in front of a promoterless GFP construct. Cells were either mock infected or infected with myxoma virus and then transfected with promoterless GFP, R-GFP, or L-GFP constructs. The L-GFP but not the R-GFP sequence was able to drive some detectable GFP expression in the absence of virus infection, but a myxoma virus coinfection greatly increased the level of expression of the L-GFP construct (Fig. 4c). From these data we conclude that the conserved sequence could act as a late promoter element for the gene 23.5L; however, other potential functions such as an involvement in viral replication or packaging cannot be excluded. The reason for the unusual conservation of this promoter sequence across four genera of poxviruses remains to be determined.
Identification of two new conserved poxvirus gene families.
The central region of the poxvirus
genome is inevitably enriched for genes that are highly conserved among
all poxviruses. In YMTV, this conserved region maps between YMTV24L and
YMTV124R. However, inspection of the genomic sequences from a number of
poxviruses revealed that the region between YMTV ORFs 27L and 29L and
between ORFs 120L and 121L were unexpectedly divergent (Fig.
5a). Analysis of the region between YMTV27L and YMTV29L identified an ORF,
designated 28.5L, which encodes a 58-amino-acid protein (Table
1). Initially we examined
the region between 27L and 29L in YLDV, where the previously assigned
28R gene is present. Analysis of the YLDV sequence revealed a clear
ortholog of 28.5L (Fig.
6a) which overlaps extensively with 28R. Based on the fact that there are
no other reported poxvirus versions of YLDV 28R in the database and
there is typically only minor overlap of poxvirus ORFs with each other,
we postulate that 28.5L represents the true yatapox virus ORF that maps
between 27L and 29L for both YMTV and YLDV and that the slightly longer
28R encoded in the opposite polarity originally annotated for YLDV
might not be expressed.
![]() View larger version (35K): [in a new window] |
FIG. 5. Alignment
of two conserved ORF clusters in a variety of poxvirus genera. (a)
Alignment of predicted ORFs from representative members from seven of
the eight poxvirus genera. Two regions of the genome are shown, the
orthologous region between YMTV027L and YMTV029L and the region between
YMTV120L and YMTV121L. Orthologous ORFs share the same color. (b)
Analysis of the region between YMTV027L/029L and YMTV120L/121L revealed
two new conserved gene families. The proposed arrangements of these
ORFs are shown, highlighting the arrangement of the two new gene
families YMTV028.5L and
YMTV120.5L.
|
![]() View larger version (35K): [in a new window] |
FIG. 6. Alignments
of the predicted YMTV28.5L and YMTV120.5L protein families. (a)
Arrangement of YMTV 28.5L orthologs in YMTV, YLDV, and myxoma virus.
(b) Alignment of orthologs of YMTV28.5L, including YLDV28.5L,
LSDV (LSDV28.5), myxoma virus (M023.5L), vaccinia virus (vvF14L), SPV
(SPV26.5), and molluscum contagiosum virus (MC022.1L). (c) Alignment of
orthologs of YMTV120.5L, including YLDV120.5L, LSDV (LSDV120.5), myxoma
virus (M119L), vaccinia virus (vvA30.5L), swinepox virus (SPV117.5),
fowlpox virus (FPV194.5L), and molluscum contagiosum virus
(MC137L).
|
We next examined the 199-bp noncoding region between YMTV ORFs 120L and 121L. A single 44-amino-acid ORF designated YMTV120.5L was identified which lacked sequence similarity to any gene in the published database. Therefore, as in the case of the gap between 27L and 29L of YMTV, we examined the sequence gap between YMTV120L and YMTV121L (Fig. 5a) and looked for other poxvirus ORFs in this conserved region. This approach yielded clear orthologs of YMTV120.5L in all poxvirus species examined (Fig. 5b and 6c). Interestingly, versions of YMTV120.5L were previously identified in myxoma virus and molluscum contagiosum virus, although originally no relationship was reported between them, presumably because the small gene size made determination of significant identity difficult. However, the position of the conserved ORF in the genomes, the sequence similarities, and the similar gene sizes all indicate that these ORFs are part of an ancestrally evolved gene cluster that is conserved across multiple poxvirus genera.
|
|
|---|
The data presented here highlight the utility of using a comparative genomic approach when analyzing viral genomes for predicted genes. One of the difficulties in whether to assign a nucleotide sequence as an annotated ORF, particularly for small ORFs of less than 150 nucleotides, is that there is no way to confirm that a predicted ORF is actually expressed until the translated protein or mRNA is detected experimentally. However, we reasoned that if a putative ORF actually encodes a protein, it would be conserved in at least some other poxvirus genus members. Therefore, we examined the tentatively assigned noncoding regions between ORFs in poxvirus genomic sequences to identify yatapoxvirus ORFs with demonstrable similarity in terms of size, sequence, and presence of contiguous orthologs. This approach identified three new yatapoxvirus gene families (23.5L, 28.5L, and 120.5L) that are clearly conserved throughout many genera of poxviruses (Table 3). These three gene families all appear to encode unique proteins with no significant similarity with any other viral or cellular proteins in the sequence database, but which are clearly conserved in most of the known poxvirus genera. With the renewed interest in variola virus, the causative agent of smallpox, it is particularly relevant to identify new families of conserved viral genes that may have important conserved roles in poxvirus replication or pathogenesis.
|
View this table: [in a new window] |
TABLE 3. Members
of three new poxvirus gene families
|
Comparing genomic sequences from different poxviruses in this fashion can provide insight into the evolutionary history of these viruses. For example, comparing the presumptive noncoding regions of both YLDV and YMTV with the same region of SPV revealed a potential pseudogene in YLDV and YMTV that had significant sequence similarity with the SPV002 gene. The presence of the same pseudogene in both YMTV and YLDV but of a functional copy of the gene in SPV implies that the pseudogene arose after the split of the suipox viruses from the yatapox viruses. In this way, we can develop an evolutionary timeline for some of the major events that differentiated members of the diverse poxvirus genera.
In addition to the identification of potential ORFs, the comparative genomic approach resulted in the unexpected identification of a 40-nucleotide stretch of YMTV sequence that was 100% conserved across members of the Yatapoxvirus, Suipoxvirus, and Capripoxvirus genera. This domain represents the most highly conserved sequence yet described among these poxviruses. Even the highly conserved concatemer resolution sequence, which is involved in the essential elements of poxvirus replication at the termini, is only 81% conserved between these species. This conserved sequence maps in the noncoding region between YMTV ORFs 23.5L and 24L. Although we demonstrated that this sequence can function as a late promoter element (Fig. 4), it is not yet clear if that is the actual function of this sequence during a viral infection. For example, the poxvirus concatemer resolution sequence can function as a poxvirus late promoter element (TAAAT) sequence (28); however, its primary role appears to be in resolving concatemers during viral replication (19). One way to test the potential function of this conserved promoter-like sequence would be to generate virus deletion mutants in any one of the virus members that contain a copy of the sequence.
The data presented here have illustrated some of the potential applications of taking a comparative approach to analyze poxvirus genomics. Through the comparison of poxvirus genomes across genera we identified three new gene families that had previously been overlooked because of their small size. In addition, conserved sequences that do not encode an ORF but that potentially play an important role in poxvirus replication were also identified. The comparative genomic analysis that we undertook was originally made possible due to the sequencing of the YMTV genome and the ability to compare its sequence to that of another relatively close species, YLDV (15). However, in theory, the comparative approach that we took could be applied to any viral family and may be particularly valuable when trying to predict whether small potential ORFs truly encode a protein.
This work was supported by the National Cancer Institute of Canada and by Viron Therapeutics, Inc.
Present
address: Department of Biology, Trent University, Peterborough, Ontario
K9L 7B8, Canada. ![]()
|
|
|---|
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2009 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»